We propose to investigate the behavior of severely and profoundly developmentally retarded (DR) children on concurrent schedules of reinforcement in order to extend Herrnstein's (1970) finding that subjects match relative rate of responding to relative rate of reinforcement. Employing Herrnstein's quantitative expressions of this relation, we plan a three-fold investigation: 1) The establishment of "matching" as a constant datum for the analysis of DR subject whose respone rate inconstancy on simple schedules of reinforcement prevents the development of an evaluative baseline; 2) An analysis of attentional deficits in terms of choice and reinforcement for competing responses (Ro); and 3) A direct measurement of reinforcer preference. These studies would be directed toward the design of prosthetic learning environments and the determination of optimal motivational conditions for developmentally retarded children.